considerable value may become the marginal bidders for many types of Government securities-'

From this point the emphasis in Mr. Murphy's paper changes, and the injurious effects of an abnormally low Government interest rate on many classes of investors is stressed. This is the second point brought out in the general summary of conclusions quoted on page 181 above, to the effect that "many holders of tax-exempt securities suffer an important loss in yield, which loss they are able to recoup only partially, or not at all, through savings in taxes." Thus, the beneficiaries of public trust funds are said to be injured by the low yield on the tax-exempt securities held in these funds. Mutual savings banks, building associations, insurance companies, and many commercial banks, are said to be investing, or trying to invest, the funds of others who are mainly persons with small incomes. If the exemption privilege were eliminated, the revenue which these institutions receive from government bonds would be increased, except as the income tax exemption which some of these groups now enjoy might later be removed. A plea is also made for the small investor, who ought to get a larger interest return on his investment in Government securities.

The magnitude of the gains which all of these classes of investors would be expected to enjoy, after the removal of tax exemption, is suggested by the statement, quoted above, that many of them now suffer an important loss in yield. This clearly implies that the interest rate on tax exempt securities has been abnormally depressed and that the removal of the exemption would produce a sufficient rise of interest rates on public securities to restore this important loss of yield. But in a later section of the paper, as will be seen, the effects of removing the tax exemption on interest rates and interest costs are minimized and made to appear of little importance. This is part of the general effort to show that interest savings will be far less than revenue losses. Just how an important loss of yield to investors can be made good without incurring an important increase of interest costs to the debtor governments is nowhere considered.

Furthermore, it is hinted that such gains as various institutions might temporarily enjoy from a rise of interest rates following the elimination of tax exemption will last only until these gains are cut into by Federal or local taxes. It is most unlikely that the States will ever tax their own pension funds, but some Federal departments believe that a Federal right to tax these State trust funds already exists.18 The third stage of the general conclusion quoted on page 181 above is that the net balance of the alleged revenue loss to government and of the investment loss to certain investors goes to the benefit of a few wealthy individual holders of such securities.

This viewpoint has already been discussed at length in the foregoing report. With respect to State and local securities, it is clear that whatever increase of interest costs may be produced by the elimination of their immunity from Federal taxation will fall upon the taxpayers who must support State and local government. This group is not mentioned in Mr. Murphy's paper, nor is the effect of the proposed tax changes upon it a matter about which any concern is expressed. As the language that is quoted on page 181 indicates, the subject is treated as if the only matter involved were the recovery from a few individuals of the benefits now derived by them from tax exemption. Were it possible to effect such recovery without producing equal or even more severe adverse effects in other directions, there would be no objection to the move from a fiscal standpoint, although the case for doing it in an orderly and definite manner by constitutional action would be as strong as ever.

However the action against the few wealthy individuals be taken, it is impossible to penalize them without also penalizing many millions of small taxpayers throughout the nation. The problem is not as narrow and one-sided as Mr. Murphy, together with many others, has assumed; it is a complicated problem which involves careful balancing of gains and losses in various directions.

Finally, the statement under review deals briefly with the effect of the removal of tax exemption on Government interest rates. The approach here is that of showing the relative unimportance of the tax-exemption privilege as a factor in market valuation, and hence of showing that the elimination of exemption privileges will cause but little readjustment in interest rates. Even so, Mr. Murphy concedes that the interest rate on long-term debt will be increased by one-fourth to one-half percent. He estimates the value of partial exemption to individuals at one-eighth percent, but does not consider the fact that all Federal bonds and notes are completely exempt when held by corporations. However, on the basis 1 Supra, p. 115.

of these estimates of the effect of the removal of tax exemption, the additional interest cost to the Federal Government on the debt volume as of June 30, 1937, produced by eliminating the exemption of Federal interest from Federal income tax, would be from $90,000,000 to $110,000,000 annually. This calculation is made by assuming one-eighth percent increase for the partially exempt debt, and an average increase of three-eighths percent for the wholly exempt debt, which produces a total increase of $90,000,000; or an increase of one-half percent on the long-term debt, which would produce the total of $110,000,000. In other words, from the purely fiscal standpoint, the Federal Government starts with a handicap of not less than $90,000,000 to $110,000,000 in increased interest cost, and the net advantage to the Treasury will consist only in the amount of revenue to be realized above such figures.

Mr. Hanes, in his statement before the special Senate committee, accepted Mr. Murphy's suggestions as to the effect of Federal interest rates, but he offered an estimate of total increased interest cost of from $19,000,000 to $50,000,000. There is a very definite and obvious discrepancy between the conceded effects on interest rates and the resulting effects on total interest cost.

The calculations of Federal interest cost in this report have assumed 20 points increase for the short-term debt, except for the bills and certificates, and 60 points for the long-term debt. These ratios of increase are in substantial agreement with the one-fourth to one-half percent suggested by Mr. Hanes. It is assumed, in this report, that the partially exempt bonds will react to the removal of exemption in a greater degree than is conceded by Mr. Hanes, because of the fact that they are now wholly exempt when held by corporations of every description.

It seems proper to suggest that the Treasury has available certain data which, if properly utilized, might go far toward settling the argument about the classes of investors who establish the interest rates on the Federal debt. These data consist, first, of the factors that are considered by Treasury officials in deciding upon the terms of a bond or note issue, and second, of the subscriptions which are made to the successive offerings. If careful minutes had been kept of the various staff conferences which precede the announcement of each offering, it should be possible to ascertain the extent to which the terms of issue had been fixed with reference to the attitude of the various classes of potential subscribers. It would be possible to discover if these conferences gave any weight to the attitude of those persons with small incomes, who are supposed by many to be the marginal buyers and whose attitude is supposed effectively to determine the interest terms of the issue. From an analysis and classification of the subscriptions and allotments, it would be possible to show who are the original purchasers of the successive issues, and with this information it would be possible to arrive at some fairly definite inferences as to the motives which appeared to govern them in making these subscriptions. With such information it would be possible to estimate more accurately the probable effects of the removal of tax exemption. Another, and even better suggestion, was offered by Dr. C. O. Hardy in the discussion of the papers given at the meeting of the American Statistical Association in Detroit on December 29, 1938. This was that in forthcoming bond issues, the offering be divided, part being wholly exempt, part only partially exempt, and part having no exemption privilege. This would provide a perfect comparison, as the only difference in the attractiveness of the bonds would be the difference in taxable status. Any market differentials that were established would supply a fairly definite clue to the current appraisal of the tax-exemption privilege. Messrs. Hanes and Murphy alike reveal a certain confusion in their discussion of the way in which the price and yield bases of the tax-exempt bonds are determined. For illustration some passages in Mr. Hanes' statement are used. In one place Mr. Hanes says:

"The reason for the small differential in interest rates, despite the high preferential value of the tax exemption privilege is that the interest saving arising from the issuance of tax-exempt securities is measured only by the value of tax-exemption to those bondholders who fall in the lowest tax brackets."

This statement obviously rests on the kind of reasoning advanced by Mr. Murphy, and dealt with in the early part of this memorandum. The fallacy as to the amount of interest saving which results from proving that the small investor is the marginal buyer by assuming that he is the marginal buyer has been pointed out above. Yet, here the small investor is made the controlling influence in establishing the yield basis.

But in another place, two pages farther on, Mr. Hanes says:

"Persons with large incomes derive much greater benefits in reduced taxes than they pay for through sacrifice of interest returns. Part of this excess benefit

falls as a burden on holders of tax-exempt securities who need them for other reasons but must pay the same premiums as do the individuals in the higher income brackets." [Italics ours.]

Here it is evident that the individuals in the high-income brackets are regarded as the principal factor in establishing the price and yield basis, and all others who want public securities must meet the terms which such individuals establish.

This inconsistency of saying that the marginal buyer is the small investor, and then that he is the large investor is so obvious and so extraordinary that no comment is needed.

The sentences of Mr. Hanes' statement which follow immediately after the second passage quoted above indicate a certain wishful thinking. These sentences

are:

"The remaining excess benefit must be paid by the general taxpayer who is called upon to make up the deficit in revenue. Neither of the burdened groups (i. e., those who accept an abnoramlly low return and the general taxpayers) is as able to bear the additional load as are the individuals in the higher income brackets who receive the benefits."

This statement implies that the individuals in the high-income brackets will continue to hold just as many of the public securities after the removal of tax exemption as they now hold. The conclusions arrived at elsewhere by Messrs. Hanes and Murphy indicate a belief that they will also hold them at a price and yield basis very little above the present levels, for both of these writers minimize the increase of interest costs while stressing greatly the prospective revenue yields. But what assurance is there that these bonds will stay in the portfolios of those in the high-income brackets once the exemption is removed? It is strongly implied by both writers that these persons attach a higher importance to the exemption than to other features. When that particular feature is removed, they would have little or no reason for carrying them longer. Hence the suggestion that they could be made to carry an additional tax load, through removal of the exemption, is a hope rather than a certainty.

If the public securities should be generally dislodged from the high-income brackets by the removal of tax exemption, where will they go and what will be the fiscal results of this shift? Naturally, they must pass into possession of those to whom the other features of a public security are supposed to be of more importance than tax exemption. Unless these groups have an indefinite capacity to absorb the securities at only such slight advance over the current interest rates as is predicted by Messrs. Hanes and Murphy, additional inducement must be provided in higher rates. Insofar as these interest rates rise, the burden on other taxpayers is increased, while at the same time the prospective revenues will diminish, for the groups that are supposed to be least concerned with tax exemption as such are the small investors, the exempt institutions, and the corporations likely to have little or no net income anyway. Little or no income tax is collected from these groups. Neither Mr. Hanes nor Mr. Murphy has given any attention to the balance of gains and losses under such a condition as this, although its emergence is to be strongly inferred from the line of analysis which they have followed.

Should the interest rates rise to a level that would again attract those with substantial incomes, the revenue will rise, but the interest costs will also have risen in the meantime.

The CHAIRMAN. Dr. Lutz, we are glad to have you here.

STATEMENT OF DR. HARLEY L. LUTZ, PROFESSOR OF PUBLIC FINANCE, PRINCETON UNIVERSITY, PRINCETON, N. J.

Dr. LUTZ. Mr. Chairman and gentlemen of the committee: The report which Mr. Tremaine has handed you as coming from me is in two main sections, as you will observe; one, an examination of the physical results of various possibilities of dealing with the taxation of public securities, and the other, an examination of some of the general governmental and social aspects of that probability.

The CHAIRMAN. As I understand it, you do not attempt to cover the legal phases in this report?

Dr. LUTZ. No, sir; not at all. In the first part of my report, dealing with the physical results, I have approached it from various points

of view-various possibilities in dealing with questions of public securities, and I have called them, for convenience, various options. Option 1 is the first and simplest case, in which the Federal Government taxes the interest of State and local securities, with no thought of reciprocity of State taxing of Federal interest.

The second option is one in which there is complete reciprocity.

The third option is one which relates to the Federal Government alone. Federal taxation of Federal interest. It is obvious that is something that can be dealt with either by itself or in combination with the other possible requirements and constitutional action.

There is no controversy between the Federal Government and the State in this connection, and, therefore, it is something that can be dealt with at any time that the Congress sees fit.

Options 1 and 2: As a matter of fact, the subject falls into two parts, because the whole question of Federal relationship may be involved as to the way to proceed, while the third option involves none, and, as a matter of fact, the whole question of reciprocity has dealt simply with this matter of State and Federal taxation. The amendment of 1922 did not couple with it Federal taxation of Federal interest. It is a question of whether there was any intention to couple with a proposed amendment today Federal taxation of Federal interest.

Now, if we do not include in an amendment, assuming we are coming to that stage, a requirement that the Federal Government tax its own interest as a condition of taxing State interest, obviously there will be no taxation of Federal interest except at the discretion of the Congress, and, whether the present Congress should do it or not, a future Congress could cover taxation of Federal interest as well.

If we should make it obligatory that the Federal Government tax its own interest as well as the State interest, then, in any future Congress, we are going to be in a very serious difficulty with respect to Federal interest because of that restriction of its policy.

Now, let us turn to a consideration of these various options, and I will outline very briefly the results I have obtained. They are shown on this chart on the wall. Small reproductions of the same chart, I believe, are in your hands.

The CHAIRMAN. I have one, but I do not think that Senator Byrd has one.

(The chart shown on p. 188 was placed before the members of the committee.)

Dr. LUTZ. Now, the first option, which is simply the Federal taxation of State and local interest with no reciprocity. Obviously, the State would get no revenue, so it becomes a tax with only Federal revenue receipts.

I think it would not be relevant, Mr. Chairman, for me to go at any length into the methods I have used in arriving at these results, but I should like to show, with respect first of all as to the cost to the State, that I have approached that from the standpoint of the probable effect upon State interest payments that would be produced by the present scheme of Federal tax rates upon the holders of State and local securities.

Now, the chief problem is to decide how much of effect will be produced by these taxes. Naturally, we had to proceed from the standpoint of opinion and inference, to a considerable extent, because,

so far, there have been no State securities subject to Federal tax, and, therefore, we have no positive evidence. The report outlines various kinds of opinion evidence that we secured by a canvass of investment houses dealing in State and local securities, a canvass of

the large insurance companies and examination of the actual market situation-market spending and also what you might call a deductive approach; that is, what would happen on the assumption the investors sought to shift the tax.